Seminario de Astrofísica en el IFIS. El próximo martes 4 de noviembre a las 14:30 hrs, Catalina Riveros, Licenciada en Astrofísica, Universidad Técnica Federico Santa María – USM, presentará el seminario:
“The halo-galaxy connection through the lens of a machine: predicting galaxy clustering”.
La presentación se realizará en la Sala 208 del Instituto de Física (IFIS), 2° piso, Edificio de Ciencias, Campus Curauma, PUCV.
Revisa el abstract de la presentación:
The large-scale linear bias encapsulates the relation between the clustering of galaxies and that of the underlying dark matter (DM) density field. A recently developed object-by-object estimator (b_inv) provides significant analytical advantages as compared to the traditional population-based methods. In this pioneering work, we employ the IllustrisTNG300 simulation, combining cutting-edge Machine Learning (ML) techniques to predict, for the first time, b_inv from DM halo and environmental properties. This framework offers a new perspective into the halo-galaxy connection. By evaluating multiple ML models, we demonstrate that deterministic methods fail to accurately reproduce this estimator, revealing the necessity of probabilistic models to capture the intrinsic stochasticity of b_inv—a critical insight given the nonlinear nature of galaxy and halo formation. Our analysis identifies the most relevant galaxy bias dependencies, and establishes a foundation for measuring individual bias in future galaxy surveys like DESI.
Extendemos la invitación a la comunidad PUCV y al público interesado.
